SUMMARY OF RESEARCH BEING UNDERTAKEN

Summary

Several years ago, Dr. Jackie Bickenbach and I wrote an article for EB Currents explaining the concept of gene therapy (see EB Currents Vol. 18, No.2, 1996). As we stated in that article, "Gene therapy is using genes to treat diseases. In the case of EB, the genes that are defective produce proteins or building blocks, which act like structural supports within a building and are necessary to maintain the overall structural integrity of the skin." Research in my own laboratory has focused on EB simplex (EBS), which is caused by mutations in genes responsible for producing the structural proteins keratins K5 and K14. Prior to testing gene therapy approaches for EBS in humans, we wanted to utilize a pre-clinical animal model to determine the safety and efficacy of these approaches. Ideally, one would like to use an animal model that mimics the human disease at the genetic level, i.e. the animal model should have the same mutation that is frequently associated with the human disease. We have recently generated a transgenic mouse model that mimics EBS at the genetic level. This mouse model has provided new insights into the molecular and cellular basis for EBS and suggested new gene therapy strategies for this disease.

The Disease Target, EBS

The four most common EBS types, include the generalized forms Koebner and Dowling-Meara, the localized form Weber–Cockayne and the Ogna variety which is found in Norwegian kindreds. The most severe form of EBS, epidermolysis bullosa herpetiformis or Dowling-Meara (EBS-DM), presents at birth with generalized blistering. Blisters occur characteristically in groups on the trunk and extremities, including palms and soles and usually heal without scarring. Blistering occurs within the epidermis and is due to lysis of basal keratinocytes in the basal layer as a result of disintegration of the keratin intermediate filament network, which normally forms a scaffold to stabilize keratinocytes. The keratin network in basal keratinocytes is composed of proteins synthesized from two different keratin genes, K5 and K14, and mutations in either gene can cause EBS. The majority of cases are due to dominant mutations; i.e. in most EBS patients, one of the two copies of their K5 or K14 genes is normal while the other one is defective due to a mutation. therefore, most EBS patients have one normal copy of their K5 or K14 keratin genes and one defective copy. Approximately 70 % of the patients with the Dowling Meara form of EBS have a mutation in exactly the same location within their K14 gene, position 125, which is referred to as a "mutational hot spot" (For additional information, see Irvine and McLean, 1999).

Development of a Ttransgenic mMouse mModel of EBS-DM

To develop a transgenic mouse model of EBS-DM, we initially isolated a DNA fragment containing the mouse K14 gene. We then engineered a point mutation into the mouse K14 gene that is equivalent to the 125 mutation found in the majority of patients with EBS-DM. The next step in generating a mouse that mimics EBS-DM in patients, requires that we replace one normal copy of the mouse K14 gene with the defective copy that we engineered to contain the 125 mutation. This is accomplished using mouse embryonic stem (ES) cells.

If you have followed the recent discussion of cloning andstem cells in the press, you probably are aware that stem cells have all of the genetic information required to generate an entire organism. So, mouse ES cells have the ability to generate all cells of an intact mouse when they are placed in the right environment. Furthermore, using techniques of molecular biology, it is relatively easy to exchange one copy of the K14 gene in mouse ES cells with a defective copy that contains the 125 mutation. This procedure is illustrated in Figure 1. ES cells are isolated from a mouse with light tan fur. These cells can be grown in culture and expanded in number. The mutated K14 gene is introduced into these cells and we select the rare cells (dark tan color) in which one normal copy of the K14 gene was replaced with the mutated copy. We were able to select for these rare genetically modified ES cells because we had inserted a drug resistance gene within the mutated copy of the K14 gene. When the selective drug is added to the cultured ES cells, the only cells that are able to survive are the ones that have incorporated the drug resistance gene, and thus the mutated copy of the K14 gene into their DNA. These genetically modified ES cells are then expanded and injected into mouse embryos at an early stage of development, prior to implantation. The embryos are isolated from mice with white fur. The ES cells containing mutated K14 will integrate into the host embryo and contribute to the formation of all tissues. The injected embryos are then surgically transferred into the uterus of a foster mother where development is allowed to progress to term. The mice that develop from the injected embryos are chimeric, that is they are composed of cells derived from the injected ES cells and the host embryo and therefore have patches of fur of both colors. By mating chimeric mice with normal mice, it is possible to determine if the injected ES cells contributed to the formation of germ cells. If this occurs, then it is possible to generate a new mouse (dark tan fur), completely derived from the ES cells and containing one normal copy of the K14 gene and one mutated copy.

The 125 mutation found in most EBS-DM patients is dominant, i.e. the keratin protein produced from the mutant copy of the K14 gene interferes with the ability of the normal K14 protein to form keratin filament networks. Therefore, if our mouse model truly mimicked the human disease, we would expect to see blisters develop on newborn mice that were derived from the ES cells that were genetically altered to contain one mutant copy of the K14 gene. Large blisters did develop on the forelimbs and chest of newborn EBS-DM mice (For details see Cao et al., 2001). In fact, the blistering was so severe that these mice died within a week after birth. Therefore, we were not able to use these mice to test gene therapy approaches for EBS.

One way to overcome this problem would be to restrict expression of the mutant copy of the K14 gene protein to a small area of the skin. Although this may seem like magic, we have now developed a mouse model that allows us to do just that, i.e. induce expression of the mutant K14 gene by applying an activator topically to the skin. Our new inducible mouse model is shown schematically in Figure 2. To prevent constitutive expression of the mutant K14 gene, we inserted a "stop sequence" into the gene prior to introduction into ES cells as illustrated in Figure 1. Since we eventually wanted to remove the stop sequence and allow expression of the mutant K14 gene, we placed LoxP sites around the stop sequence. This DNA construct was introduced into ES cells and generated the mouse shown in the top right corner of Fig. 2. Note that this mouse does not express mutant K14 protein.

LoxP sites are very specific sequences of DNA recognized by an enzyme called Cre recombinase. Just think of Cre recombinase as molecular scissors that will only cut DNA at LoxP sites. We modified the Cre recombinase to make it inducible, i.e. it would only cut LoxP sites after the addition of an activatorit was induced by an activator. In order to express the inducible Cre in the basal layer of the epidermis, where the K14 gene is normally expressed, we made a DNA vector that uses regulatory sequences of the K14 gene (the K14 promoter) to control expression of the inducible Cre. This vector was used to generate the mouse shown in the top left corner of Fig. 2.

By mating the mouse on the left with the mouse on the right, we generate a mouse with two genetic changes, it contains the mutant K14 gene with the stop sequence and it is also expressing a form of Cre recombinase in basal epidermal cells that is not active the inducible Cre recombinase. When an activator is topically applied to a small area of the skin of this mouse, Cre recombinase is activated in the basal layer of the epidermis removing the stop sequence should be removed by Cre, and this should results in activation of the mutant K14 gene. After several days of treatment of one paw with the activator, we observed a large fluid-filled blister (Fig. 3). Histological analysis revealed that blistering occurred in the basal layer of the epidermis, as expected. The blistered areas healed without scarring after seven days without treatment with the activator. No more blisters formed without further exposure to the activator. An explanation for the lack of recurrent blisters is given below.

Evidence Supporting the Hypothesis that Induced EBS Blisters are Repaired by the Migration of Normal Epidermal Stem Cells.

The epidermis is constantly regenerated by stem cells that reside within the basal layer. It is possible that the stop sequence was only excised from the mutant K14 gene in the non-stem cell population of basal keratinocytes (transit amplifying cells) in the blistered area. Although blisters were induced through cytolysis of these cells, the stem cells were not targetedaffected, i.e., they retained the stop sequence and repopulated the area after seven days. However, this is not likely the case based on our observations where the same inducible system was used to develop a mouse model for epidermolytic hyperkeratosis (EHK) (See Arin et al., 2001 for more information). EHK is very similar to EBS except that blisters develop in the suprabasal keratinocytes (the upper layers of the epidermis) due to a dominant mutation in either keratin K1 or K10, while the basal cells of the epidermis (including the stem cells) remain normal. Unlike the induced EBS blisters, which disappear, the induced EHK lesions persist (Fig. 4), to date for over 1 year. Because mouse epidermis is completely replaced by the normal shedding process every 8-10 days, this result clearly suggests that epidermal stem cells are targeted by our inducible system.

The schematic shown in Figure 5 illustrates what we have observed in these inducible models. In the inducible EBS model, focal activation of the mutant K14 gene occurs in stem cells, and since the mutant K14 protein is expressed in these cells, they are fragile. With time, neighboring normal stem cells (where the stop sequence is still in place) migrate into the area and displace the lysed EBS stem cells. In the inducible EHK model, removal of the stop sequence from the mutant K10 gene also occurs in stem cells. However, the mutant K10 protein is not expressed in stem cells, only in the progeny of stem cells after they have differentiated and moved into the suprabasal layers. Therefore, there is no selection against epidermal stem cells with K10 mutations and these stem cells continue to give rise to defective differentiated progeny for the life of the mouse.

Inducible Mouse Models for EBS and EHK Define the Role of Stem Cells in Mosaic Skin Disorders.

Patients with mosaic forms of EHK, i.e. patients with mostly normal skin that have patches of skin that exhibit EHK have been described. A molecular analysis of these patients revealed the presence of K10 mutations in the affected areas, but not in the normal skin [see Paller et al., 1994]. Mosaic forms of EHK probably occur as a result of mutations in K10 that arise after the embryo begins to develop. and the distribution of affected epidermis reflects migration and proliferation of mutant epidermal stem cells. Therefore, the developing embryo will have normal cells and a small percentage of cells with mutant K10, which can contribute to the formation of the skin. Our analysis of the inducible EHK mouse model clearly suggests that in EHK, mutant epidermal stem cells can exist side by side with wild-type stem cells. Since the mutant K10 allele protein is not expressed in the basal layer (containing the stem cells), there is no selection against mutant EHK epidermal stem cells with K10 mutations. In contrast, a selection process takes place against defective epidermal stem cells in the inducible mouse model for EBS, when the mutant K14 allele is focally activated in the basal layer. The mutant stem cells are replaced with normal stem cells migrating from untreated areas of the skin. Since there are no reports of patients with mosaic forms of EBS, our results indicate that a lack of selective pressure against K10 mutations could explain why mosaic forms exist for EHK.

Implications for Gene Therapy

The inducible transgenic mouse models for EBS and EHK have not only provided new insights into the molecular and cellular basis of mosaic skin disorders, but also suggested new strategies for somatic gene therapy for keratin disorders. For example, the healing of induced blisters and the absence of recurrent lesions in the EBS model raises the possibility that defective EBS stem cells may be replaced by non-defective stem cells. One approach to test this hypothesis would be to isolate EBS stem cells, genetically modify them to eliminate expression of the mutant K14 gene and return these corrected stem cells into blistered areas to see if they have a selective growth advantage. We are currently using our mouse model for EBS to test these strategiesthis strategy.

Acknowledgements:

This research was supported by grants from DebRA of America and the National Institutes of Health (AR 62228 and HD 25479). The transgenic mouse models described in this article were generated by two very talented postdoctoral fellows: Dr. Tongyu Cao, who created the EBS model, is currently in the Pacific Biomedical Research Center, University of Hawaii-Manoa; Dr. Meral Arin, who created the EHK model, is currently in the Department of Dermatology, University of Cologne, Cologne, Germany. I would like to give special thanks to Maranke Koster for comments on the content of this article, as well as for her assistance in making the figures.

References:

Arin, M.J., Longley, M.A., Wang, X.J., and Roop, D.R. (2001): Focal activation of a mutant allele defines the role of stem cells in mosaic skin disorders. J. Cell Biol. 152:645-650.

Cao, T., Longley, M.A., Wang, X.J., and Roop, D.R. (2001): An inducible mouse model for epidermolysis bullosa simplex: Implications for gene therapy. J. Cell Biol. 152:651-656.

Irvine, A.D. and, McLean, W.H. (1999): Human keratin diseases: the increasing spectrum of disease and subtlety of the phenotype-genotype correlation. Br. J. Dermatol. 140:815-828.

Paller, A.S., Syder A.J., Chan, Y.M., Yu, Q.C., Hutton, E., Tadini, G., and Fuchs, E. (1994): Genetic and clinical mosaicism in a type of epidermal nevus. N. Engl. J. Med. 331:1408-1415.

Figure Legends:

Figure 1. A schematic illustrating how transgenic mice are made. Embryonic stem (ES) cells are isolated from a mouse with light tan fur and expanded in culture. The mutant K14 DNA, containing a drug resistance gene, is introduced into the ES cells. Following exposure to the selective drug, only ES cells containing the mutant K14 DNA will survive. These cells are injected into early mouse embryos isolated from mice with white fur. The injected embryos are then transferred into foster mothers and allowed to develop to term. Chimeric mice, derived from a mixture of ES cells and cells of the host embryo, have fur of both colors. These chimeras are mated with normal mice to produce mice that were completely derived from the mutated ES cells (dark tan fur), and contain one normal copy of the K14 gene and one mutated copy.

Figure 2. A schematic illustrating how the inducible EBS mouse model was generated. The mouse shown in the upper right corner was generated exactly as shown in Fig. 1, except that a "stop sequence" was placed in the mutant K14 gene to prevent expression. The mouse in the upper left corner was generated to express an inducible Cre recombinase in the basal layer of the epidermis. Cre recombinase recognizes very specific sequences of DNA called LoxP sites, which were placed around the stop sequence. Mating the mice shown in the top panel produces the mouse in the bottom panel that contains both the mutant K14 gene with the stop sequence and the inducible Cre. Topical application of an activator to a small area of the skin results in removal of the stop sequence and activation of the mutant K14 gene.

Figure 3. Focal induction of blisters in the inducible EBS mouse model. The left panel shows the induction of a blister two days following treatment with the activator on one paw. Ten days following blister induction (middle panel). Six months after blister formation the paw appears normal and no additional blisters develop (right panel).

Figure 4. Focal induction of blisters in the inducible EHK mouse model. The left panel shows blister formation on the forelimbs and chest four days after treatment. Unlike EBS induced blisters, EHK blisters persist after treatment with the activator is stopped. Shown in the right panel is a mouse at three months following blister induction. However, persistent blistering has now been observed for over one year.

Figure 5. Schematic illustrating why blisters persist in the EHK model but disappear in the EBS model. In the EBS model, the mutant K14 gene is not only activated in epidermal stem cells, but also expressed in these cells. Therefore, these cells are fragile and are replaced by normal stem cells migrating in from the surrounding non-treated area. In EHK, the mutant K10 gene is also activated in stem cells, but it is only expressed in the differentiated progeny of these cells in the suprabasal layers of the epidermis. Therefore, there is no selective pressure on EHK stem cells and they persist side by side with normal stem cells for the life of the mouse.